Switch cabinet fault classification method based on semi-supervised learning
A semi-supervised learning and fault classification technology, applied in the fields of instruments, character and pattern recognition, computer parts, etc., it can solve the problems of high dimension of monitoring features, difficulty in fault monitoring of each module of the switch cabinet, and increased difficulty in fault classification.
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0057] figure 1 It is a flow chart of the switchgear fault classification method based on semi-supervised learning in Embodiment 1 of the present invention.
[0058] Such as figure 1 As shown, the switchgear fault classification method based on semi-supervised learning in this embodiment is used to classify the fault types of the switchgear, including the following steps:
[0059] Step 1. Obtain sample data of the fault type of the switchgear, which includes labeled sample data with known fault types and unlabeled sample data with unknown fault types.
[0060] Table 1 Monitoring characteristic quantity of switchgear
[0061]
[0062] In Table 1: FS1 and FS2 reflect the working environment of the busbar; FS3-FS7 are the characteristic quantities of electrical parameters, reflecting the internal and external system faults; FS8-FS9 reflect the partial discharge of the switchgear; FS11 reflects the temperature change caused by partial discharge, etc.; FS12 reflects The break...
Embodiment 2
[0125] This embodiment selects the sample data of the fault type of a known switchgear in a power grid as a data sample, randomly selects a part as labeled samples, and the rest of the samples as unlabeled samples, and adopts the switchgear fault based on semi-supervised learning in Embodiment 1 The classification method is used to classify the fault types of the switchgear. The specific classification process is as follows:
[0126] First, preprocess the sample data:
[0127]
[0128] Accordingly, a labeled sample data set X of the switchgear is established 1 ={x 1 ,x 2 ,...,x m} and unlabeled sample dataset X 2 ={x m ,x m+1 ,...,x n}.
[0129] For the labeled sample data, the Laplacian score method is used for feature selection, and the training set S is obtained. 1 ={s 1 ,s 2 ,...,s m}. Laplace score formula:
[0130]
[0131] Train an initial classifier of fault types:
[0132]
[0133] Determine the sample point position of the switchgear accordin...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More 


